YIELD¶

YIELD defines the output of an nGQL query.

YIELD can lead a clause or a statement:

• A YIELD clause works in nGQL statements such as GO, FETCH, or LOOKUP.
• A YIELD statement works in a composite query or independently.

OpenCypher Compatibility¶

This topic applies to nGQL extensions only. For the openCypher syntax, use RETURN.

YIELD has different functions in openCypher and nGQL.

• In openCypher, YIELD is used in the CALL[…YIELD] clause to specify the output of the procedure call.

  NOTE: NGQL does not support CALL[…YIELD] yet.

• In nGQL, YIELD works like RETURN in openCypher.

YIELD clauses¶

Syntax¶

YIELD [DISTINCT] <col> [AS <alias>] [, <col> [AS <alias>] ...]

The syntax is described as follows.

Use a YIELD clause in a statement¶

• Use YIELD with GO:

  nebula> GO FROM "player100" OVER follow \
          YIELD $$.player.name AS Friend, $$.player.age AS Age;
  +-----------------+-----+
  | Friend          | Age |
  +-----------------+-----+
  | "Tony Parker"   | 36  |
  +-----------------+-----+
  | "Manu Ginobili" | 41  |
  +-----------------+-----+
  Got 2 rows (time spent 3378/4030 us)
  

• Use YIELD with FETCH:

  nebula> FETCH PROP ON player "player100" \
          YIELD player.name;
  +-------------+--------------+
  | VertexID    | player.name  |
  +-------------+--------------+
  | "player100" | "Tim Duncan" |
  +-------------+--------------+
  Got 1 rows (time spent 2933/5931 us)
  

• Use YIELD with LOOKUP:

  nebula> LOOKUP ON player WHERE player.name == "Tony Parker" \
          YIELD player.name, player.age;
  =======================================
  | VertexID | player.name | player.age |
  =======================================
  | 101      | Tony Parker | 36         |
  ---------------------------------------
  Got 1 rows (time spent 2963/3778 us)
  

YIELD Statements¶

Syntax¶

YIELD [DISTINCT] <col> [AS <alias>] [, <col> [AS <alias>] ...]
[WHERE <conditions>]

The syntax is described as follows.

Use a YIELD statement in a composite query¶

In a composite query, a YIELD statement accepts, filters, and reforms the result set of the preceding statement, and then outputs it.

The following query finds the players that "player100" follows and calculates their average age.

nebula> GO FROM "player100" OVER follow \
        YIELD follow._dst AS ID | \
        FETCH PROP ON player $-.ID \
        YIELD player.age AS Age | \
        YIELD AVG($-.Age) as Avg_age, count(*)as Num_friends;
+---------+-------------+
| Avg_age | Num_friends |
+---------+-------------+
| 38.5    | 2           |
+---------+-------------+
Got 1 rows (time spent 1846/2426 us)

The following query finds the players that "player101" follows and the follow degrees are greater than 90.

nebula> $var1 = GO FROM "player101" OVER follow \
        YIELD follow.degree AS Degree, follow._dst as ID; \
        YIELD $var1.ID AS ID \
        WHERE $var1.Degree > 90;
+-------------+
| ID          |
+-------------+
| "player100" |
+-------------+
| "player125" |
+-------------+
Got 2 rows (time spent 891/1411 us)

Use a standalone YIELD statement¶

A YIELD statement can calculate a valid expression and output the result.

nebula> YIELD rand32(1, 6);
+-------------+
| rand32(1,6) |
+-------------+
| 3           |
+-------------+
Got 1 rows (time spent 144/615 us)

nebula> YIELD "Hel" + "\tlo" AS string1, ", World!" AS string2;
+-------------+------------+
| string1     | string2    |
+-------------+------------+
| "Hel    lo" | ", World!" |
+-------------+------------+
Got 1 rows (time spent 154/692 us)

nebula> YIELD hash("Tim") % 100;
+-----------------+
| (hash(Tim)%100) |
+-----------------+
| 42              |
+-----------------+
Got 1 rows (time spent 164/820 us)

nebula> YIELD \
      CASE 2+3 \
      WHEN 4 THEN 0 \
      WHEN 5 THEN 1 \
      ELSE -1 \
      END \
      AS result;
+--------+
| result |
+--------+
| 1      |
+--------+
Got 1 rows (time spent 204/935 us)